Mitigating Included - and Omitted - Variable Bias in Estimates of Disparate Impact
In studies of discrimination, one popular strategy is to estimate disparities after adjusting for observed covariates, typically with a regression model.
Research
Slavery, Politics, and Causality
How has slavery shaped the politics of the United States over the last century and a half since emancipation?
A Radical Reckoning with Cultural Devastation and Its Aftermath: Reflections on Wub-e-ke-niew’s We Have the Right to Exist
Wub-e-ke-niew’s enormously unsettling book We Have the Right to Exist presents a version of indigenous philosophical thought as an alternative way of being human in the world that creates profound ins
The Measure and Mismeasure of Fairness
The field of fair machine learning aims to ensure that decisions guided by algorithms are equitable. Over the last decade, several formal, mathematical definitions of fairness have gained prominence.
Cancel culture: Heterodox self-censorship or the curious case of the dog-which-didn’t-bark
There is widespread concern that academic freedom is threatened by growing demands for intellectual conformity and attempted censorship from intolerant zealots involving ‘woke wars’ and a ‘cancel cult
Russia’s Willing Collaborators: Ukraine Needs a Measured Lustration Policy to Strengthen Security and Rebuild Democracy
As Ukrainian forces aim at a major counteroffensive against Russia, the issue of collaborators, especially in Russian-occupied areas, has drawn renewed attention.
Beyond the Ballot Box: A Conversation About Democracy and Policing in the United States
Political scientist Hakeem Jefferson (Stanford University) facilitated a discussion about race, policing, and the state of American democracy with fellow political scientists Cathy J.
The Birth of a Nation: Media and Racial Hate
This paper documents the impact of popular media on racial hate by examining the first American blockbuster: 1915's The Birth of a Nation, a fictional portrayal of the KKK's founding rife with racist
Risk Scores, Label Bias, and Everything but the Kitchen Sink
In designing risk assessment algorithms, many scholars promote a “kitchen sink” approach, reasoning that more information yields more accurate predictions.
Automated Court Date Reminders Reduce Warrants for Arrest: Evidence from a Text Messaging Experiment
Millions of people in the U.S. every year are required to attend mandatory court dates as their cases proceed through the criminal legal system.
Get smart & reliable public policy insights right in your inbox.